a

r
X
i
v
:
g
r
-
q
c
/
9
2
1
0
0
0
6
v
2


1
6

A
u
g

2
0
0
4
THE QUANTUM MECHANICS OF CLOSED SYSTEMS

James B. Hartle

Department of Physics, University of California

Santa Barbara, CA 93106-9530 USA
A pedagogical introduction is given to the quantum mechanics of closed systems, most gener-
ally the universe as a whole. Quantum mechanics aims at predicting the probabilities of alternative
coarse-grained time histories of a closed system. Not every set of alternative coarse-grained histories
that can be described may be consistently assigned probabilities because of quantum mechanical
interference between individual histories of the set. In Copenhagen quantum mechanics, prob-
abilities can be assigned to histories of a subsystem that have been measured. In the quantum
mechanics of closed systems, containing both observer and observed, probabilities are assigned to
those sets of alternative histories for which there is negligible interference between individual his-
tories as a consequence of the systems initial condition and dynamics. Such sets of histories are
said to decohere. We dene decoherence for closed systems in the simplied case when quantum
gravity can be neglected and the initial state is pure. Typical mechanisms of decoherence that are
widespread in our universe are illustrated.
Copenhagen quantum mechanics is an approximation to the more general quantum framework of
closed subsystems. It is appropriate when there is an approximately isolated subsystem that is a
participant in a measurement situation in which (among other things) the decoherence of alternative
registrations of the apparatus can be idealized as exact.
Since the quantum mechanics of closed systems does not posit the existence of the quasiclassical
realm of everyday experience, the domain of the approximate aplicability of classical physics must
be explained. We describe how a quasiclassical realm described by averages of densities of approxi-
mately conserved quantities could be an emergent feature of an initial condition of the universe that
implies the approximate classical behavior of spacetime on accessible scales.
I. INTRODUCTION
It is an inescapable inference from the physics of the
last sixty years that we live in a quantum mechanical
universe a world in which the basic laws of physics
conform to that framework for prediction we call quan-
tum mechanics. We perhaps have little evidence of pecu-
liarly quantum mechanical phenomena on large and even
familiar scales, but there is no evidence that the phe-
nomena that we do see cannot be described in quantum
mechanical terms and explained by quantum mechanical
laws. If this inference is correct, then there must be a
description of the universe as a whole and everything in
it in quantum mechanical terms. The nature of this de-
scription and its observable consequences are the subject
of quantum cosmology
Our observations of the present universe on the largest
scales are crude and a classical description of them is
entirely adequate. Providing a quantum mechanical de-
scription of these observations alone might be an inter-
esting intellectual challenge, but it would be unlikely to
yield testable predictions diering from those of classical
physics. Today, however, we have a more ambitious aim.
We aim, in quantum cosmology, to provide a theory of

A slightly shortened version of an article that appeared in

the festschrift for C.W. Misner, ed. by B.L. Hu, M.P. Ryan
and C.V. Vishveshwara, Cambridge University Press, Cambridge
(1993).

Electronic address: hartle@physics.ucsb.edu

the initial condition of the universe which will predict
testable correlations among observations today. There
are no realistic predictions of any kind that do not depend
on this initial condition, if only very weakly. Predictions
of certain observations may be testably sensitive to its
details. These include the large scale homogeneity and
isotropy of the universe, its approximate spatial atness,
the spectrum of density uctuations that produced the
galaxies, the homogeneity of the thermodynamic arrow
of time, and the existence of classical spacetime. Re-
cently, there has been speculation that even the coupling
constants of the eective interactions of the elementary
particles at accessible energy scales may be probabilisti-
cally distributed with a distribution which may depend,
in part, on the initial condition of the universe [4, 5, 6].
It is for such reasons that the search for a theory of the
initial condition of the universe is just as necessary and
just as fundamental as the search for a theory of the dy-
namics of the elementary particles. They may even be
the same searches.
The physics of the very early universe is likely to be
quantum mechanical in an essential way. The singular-
ity theorems of classical general relativity suggest that
an early era preceded ours in which even the geometry
of spacetime exhibited signicant quantum uctuations.
It is for a theory of the initial condition that describes
this era, and all later ones, that we need to spell out
how to apply quantum mechanics to cosmology. Recent
years have seen much promising progress in the search
for a theory of the quantum initial condition. However,
2
it is not my purpose to review these developments here.
1
Rather, I shall argue that this somewhat obscure branch
of astrophysics may have implications for the formulation
and interpretation of quantum mechanics on day-to-day
scales. My thesis will be that by looking at the universe
as a whole one is led to an understanding of quantum
mechanics which claries many of the long standing in-
terpretative diculties of the subject.
The Copenhagen frameworks for quantum mechanics,
as they were formulated in the 30s and 40s and as they
exist in most textbooks today, are inadequate for quan-
tum cosmology. Characteristically these formulations as-
sumed, as external to the framework of wave function
and Schrodinger equation, the quasiclassical realm we see
all about us. Bohr [8] spoke of phenomena which could
be alternatively described in classical language. In their
classic text, Landau and Lifschitz [9] formulated quan-
tum mechanics in terms of a separate classical physics.
Heisenberg and others stressed the central role of an
external, essentially classical, observer.
2
Characteristi-
cally, these formulations assumed a possible division of
the world into obsever and observed, assumed that
measurements are the primary focus of scientic state-
ments and, in eect, posited the existence of an external
quasiclassical realm. However, in a theory of the whole
thing there can be no fundamental division into observer
and observed. Measurements and observers cannot be
fundamental notions in a theory that seeks to describe
the early universe when neither existed. In a basic formu-
lation of quantum mechanics there is no reason in general
for there to be any variables that exhibit classical behav-
ior in all circumstances. Copenhagen quantum mechan-
ics thus needs to be generalized to provide a quantum
framework for cosmology.
In a generalization of quantum mechanics which does
not posit the existence of a quasiclassical realm, the do-
main of applicability of classical physics must be ex-
plained. For a quantum mechanical system to exhibit
classical behavior there must be some restriction on its
state and some coarseness in how it is described. This is
clearly illustrated in the quantum mechanics of a single
particle. Ehrenfests theorem shows that generally
M
d
2
x
dt
2
=
_

V
x
_
. (1.1)
However, only for special states, typically narrow wave
packets, will this become an equation of motion for x
of the form
M
d
2
x
dt
2
=
V (x)
x
. (1.2)
For such special states, successive observations of posi-
tion in time will exhibit the classical correlations pre-
dicted by the equation of motion (1.2) provided that these
1
For a recent review of quantum cosmology see [7].
2
For a clear statement of this point of view, see [10].
observations are coarse enough so that the properties of
the state which allow (1.2) to replace the general relation
(1.1) are not aected by these observations. An exact de-
termination of position, for example, would yield a com-
pletely delocalized wave packet an instant later and (1.2)
would no longer be a good approximation to (1.1). Thus,
even for large systems, and in particular for the universe
as a whole, we can expect classical behavior only for cer-
tain initial states and then only when a suciently coarse
grained description is used.
If classical behavior is in general a consequence only of
a certain class of states in quantum mechanics, then, as a
particular case, we can expect to have classical spacetime
only for certain states in quantum gravity. The classical
spacetime geometry we see all about us in the late uni-
verse is not property of every state in a theory where
geometry uctuates quantum mechanically. Rather, it
is traceable fundamentally to restrictions on the initial
condition. Such restrictions are likely to be generous in
that, as in the single particle case, many dierent states
will exhibit classical features. The existence of classical
spacetime and the applicability of classical physics are
thus not likely to be very restrictive conditions on con-
structing a theory of the initial condition.
It was Everett who, in 1957, rst suggested how to gen-
eralize the Copenhagen frameworks so as to apply quan-
tum mechanics to cosmology.
3
Everetts idea was to take
quantum mechanics seriously and apply it to the universe
as a whole. He showed how an observer could be consid-
ered part of this system and how its activities mea-
suring, recording, calculating probabilities, etc. could
be described within quantum mechanics. Yet the Everett
analysis was not complete. It did not adequately describe
within quantum mechanics the origin of the quasiclas-
sical realm of familiar experience
4
nor, in an observer
independent way, the meaning of the branching that
replaced the notion of measurement. It did not distin-
guish from among the vast number of choices of quantum
mechanical observables that are in principle available to
an observer, the particular choices that, in fact, describe
the quasiclassical relm.
In this essay, I will describe joint work with Murray
Gell-Mann [2, 3] which aims at a coherent formulation
of quantum mechanics for the universe as a whole that
is a framework to explain rather than posit the quasi-
classical realm of everyday experience. It is an attempt
at an extension, clarication, and completion of the Ev-
erett interpretation. It builds on many aspects of the,
so called post-Everett development, especially the work
of Zeh [13], Zurek [14, 15], and Joos and Zeh [16]. At
important points it coincides with the, independent, ear-
3
The original reference is [11]. For a useful collection of reprints
see [12].
4
In our earlier work e.g. [2] this was called a quasiclassical do-
main, but this risked confusion with the use of the word do-
main in condensed matter physics.
3
lier work of Bob Griths [17] and Roland Omn`es (e.g. as
reviewed in [18]).
Our work is not complete, but I hope to sketch how it
might become so. It is by now a very long story but I will
try to describe the important parts in simplied terms.
II. PROBABILITIES IN GENERAL AND
PROBABILITIES IN QUANTUM MECHANICS
Even apart from quantum mechanics, there is no cer-
tainty in this world and therefore physics deals in proba-
bilities. It deals most generally with the probabilities for
alternative time histories of the universe. From these,
conditional probabilities can be constructed that are ap-
propriate when some features about our specic history
are known and further ones are to be predicted.
To understand what probabilities mean for a single
closed system, it is best to understand how they are used.
We deal, rst of all, with probabilities for single events
of the single system. When these probabilities become
suciently close to zero or one there is a denite pre-
diction on which we may act. How suciently close to
0 or 1 the probabilities must be depends on the circum-
stances in which they are applied. There is no certainty
that the sun will come up tomorrow at the time printed
in our daily newspapers. The sun may be destroyed by
a neutron star now racing across the galaxy at near light
speed. The earths rotation rate could undergo a quan-
tum uctuation. An error could have been made in the
computer that extrapolates the motion of the earth. The
printer could have made a mistake in setting the type.
Our eyes may deceive us in reading the time. Yet, we
watch the sunrise at the appointed time because we com-
pute, however imperfectly, that the probability of these
things happening is suciently low.
Various strategies can be employed to identify situa-
tions where probabilities are near zero or one. Acquiring
information and considering the conditional probabilities
based on it is one such strategy. Current theories of the
initial condition of the universe predict almost no proba-
bilities near zero or one without further conditions. The
no boundary wave function of the universe, for exam-
ple, does not predict the present position of the sun on
the sky. However, it will predict that the conditional
probability for the sun to be at the position predicted
by classical celestial mechanics given a few previous po-
sitions is a number very near unity.
Another strategy to isolate probabilities near 0 or 1 is
to consider ensembles of repeated observations of identi-
cal subsystems in the closed system. There are no gen-
uinely innite ensembles in the world so we are necessar-
ily concerned with the probabilities for deviations of the
behavior of a nite ensemble from the expected behavior
of an innite one. These are probabilities for a single
feature (the deviation) of a single system (the whole en-
semble).
The existence of large ensembles of repeated observa-
tions in identical circumstances and their ubiquity in lab-
oratory science should not, therefore, obscure the fact
that in the last analysis physics must predict probabili-
ties for the single system that is the ensemble as a whole.
Whether it is the probability of a successful marriage, the
probability of the present galaxy-galaxy correlation func-
tion, or the probability of the uctuations in an ensemble
of repeated observations, we must deal with the prob-
abilities of single events in single systems. In geology,
astronomy, history, and cosmology, most predictions of
interest have this character. The goal of physical theory
is, therefore, most generally to predict the probabilities
of histories of single events of a single system.
Probabilities need be assigned to histories by physi-
cal theory only up to the accuracy they are used. Two
theories that predict probabilities for the sun not rising
tomorrow at its classically calculated time that are both
well beneath the standard on which we act are equiva-
lent for all practical purposes as far as this prediction
is concerned. It is often convenient, therefore, to deal
with approximate probabilities which satisfy the rules of
probability theory up to the standard they are used.
The characteristic feature of a quantum mechanical
theory is that not every set of alternative histories that
may be described can be assigned probabilities. Nowhere
is this more clearly illustrated than in the two slit experi-
ment illustrated in Figure 1. In the usual Copenhagen
discussion if we have not measured which of the two slits
the electron passed through on its way to being detected
at the screen, then we are not permitted to assign prob-
abilities to these alternative histories. It would be in-
consistent to do so since the correct probability sum rule
would not be satised. Because of interference, the prob-
ability to arrive at point y on the screen is not the sum of
the probabilities to arrive at y going through the upper
or lower slit:
p(y) = p
U
(y) +p
L
(y) (2.1)
because in quantum theory probabilities are squares of
amplitudes and
|
L
(y) +
U
(y)|
2
= |
L
(y)|
2
+ |
U
(y)|
2
. (2.2)
If we have measured which slit the electron went
through, then the interference is destroyed, the sum rule
obeyed, and we can meaningfully assign probabilities to
these alternative histories.
A rule is thus needed in quantum theory to determine
which sets of alternative histories may be assigned prob-
abilities and which may not. In Copenhagen quantum
mechanics, the rule is that probabilities are assigned to
histories of alternatives of a subsystem that are measured
and not in general otherwise.
4
2

U
L
y
y
FIG. 1: The two-slit experiment. An electron gun at left
emits an electron traveling towards a screen with two slits, its
progress in space recapitulating its evolution in time. When
precise detections are made of an ensemble of such electrons
at the screen it is not possible, because of interference, to as-
sign a probability to the alternatives of whether an individual
electron went through the upper slit or the lower slit. How-
ever, if the electron interacts with apparatus that measures
which slit it passed through, then these alternatives decohere
and probabilities can be assigned.
III. PROBABILITIES FOR A TIME SEQUENCE
OF MEASUREMENTS
To establish some notation, let us review in more detail
the usual rules for the probabilities of time sequences
of ideal measurements of subsystem using the two-slit
experiment of Figure 1 as an example.
Alternatives for the electron are represented by pro-
jection operators in its Hilbert space. Thus, in the two
slit experiment, the alternative that the electron passed
through the lower slit is represented by the projection
operator
P
U
=
s
_
U
d
3
x|x, sx, s| (3.1)
where |x, s is a localized state of the electron with spin
component s, and the integral is over a volume around
the upper slit. There is a similar projection operator
P
L
for the alternative that the electron goes through the
lower slit. These are exclusive alternatives and they are
exhaustive. These properties, as well as the requirements
of being projections, are represented by the relations
P
L
P
U
= 0 , P
U
+P
L
= 1, P
2
L
= P
L
, P
2
U
= P
U
.
(3.2)
There is a similarly dened set of projection operators
{P
y
} representing the alternative positions of arrival at
the screen.
We can now state the rule for the joint probability
that the electron initially in a state |(t
0
) at t = t
0
is
determined by an ideal measurement at time t
1
to have
passed through the upper slit and measured at time t
2
to arrive at point y on the screen. If one likes, one can
imagine the case in which the electron is in a narrow
wave packet in the horizontal direction with a velocity
dened as sharply as possible consistent with the uncer-
tainty principle. The joint probability is negligible unless
t
1
and t
2
correspond to the times of ight to the slits and
to the screen respectively.
The rst step in calculating the joint probability is to
evolve the state of the electron to the time t
1
of the rst
measurement

(t
1
)
_
= e
iH(t1t0)/

(t
0
)
_
. (3.3)
The probability that the outcome of the measurement at
time t
1
is that the electron passed through the upper slit
is:
(Probability of U) =
_
_
P
U

(t
1
)
__
_
2
(3.4)
where denotes the norm of a vector in the electrons
Hilbert space. If the outcome was the upper slit, and
the measurement was an ideal one, that disturbed the
electron as little as possible in making its determination,
then after the measurement the state vector is reduced
to
P
U
|(t
1
)
P
U
|(t
1
)
. (3.5)
This is evolved to the time of the next measurement
|(t
2
) = e
iH(t2t1)/
P
U
|(t
1
)
P
U
|(t
1
)
. (3.6)
The probability of being detected at point y on the screen
at time t
2
given that the electron passed through the
upper slit is
(Probability of y given U) = P
y
|(t
2
)
2
. (3.7)
The joint probability that the electron is measured to
have gone through the upper slit and is detected at y
is the product of the conditional probability (3.7) with
the probability (3.4) that the electron passed through U.
The latter factor cancels the denominator in (3.6) so that
combining all of the above equations in this section, we
have
(Probability of y and U)
=
_
_
_P
y
e
iH(t2t1)/
P
U
e
iH(t1t0)/

(t
0
)
_
_
_
_
2
. (3.8)
With Heisenberg picture projections this takes the even
simpler form
(Probability of y and U) =
_
_
P
y
(t
2
)P
U
(t
1
)

(t
0
)
_
_
2
.
(3.9)
where, for example,
P
U
(t) = e
iHt/
P
U
e
iHt/
. (3.10)
The formula (3.9) is a compact and unied expression
of the two laws of evolution that characterize the quan-
tum mechanics of measured subsystems unitary evolu-
tion in between measurements and reduction of the wave
5
packet at a measurement.
5
The important thing to re-
member about the expression (3.9) is that everything in
it projections, state vectors, Hamiltonian refer to
the Hilbert space of a subsystem, in this example the
Hilbert space of the electron that is measured.
In Copenhagen quantum mechanics, it is measure-
ment that determines which histories of a subsystem
can be assigned probabilities and formulae like (3.9)
that determine what these probabilities are. We can-
not have such rules in the quantum mechanics of closed
systems. There is no fundamental division of a closed
system into measured subsystem and measuring appara-
tus. There is no fundamental reason for the closed sys-
tem to contain classically behaving measuring apparatus
in all circumstances. In particular, in the early universe
none of these concepts seem relevant. We need a more
observer-independent, measurement-independent, quasi-
classical realm-independent rule for which histories of a
closed system can be assigned probabilities and what
these probabilities are. The next section describes this
rule.
IV. POST-EVERETT QUANTUM MECHANICS
To describe the rules of post-Everett quantum me-
chanics, I shall make a simplifying assumption. I shall
neglect gross quantum uctuations in the geometry of
spacetime, and assume a xed background spacetime ge-
ometry which supplies a denite meaning to the notion
of time. This is an excellent approximation on accessible
scales for times later than 10
43
sec after the big bang.
The familiar apparatus of Hilbert space, states, Hamil-
tonian, and other operators may then be applied to pro-
cess of prediction. Indeed, in this context the quantum
mechanics of cosmology is in no way distinguished from
the quantum mechanics of a large isolated box, perhaps
expanding, but containing both the observed and its ob-
servers (if any).
A set of alternative histories for a closed system is spec-
ied by giving exhaustive sets of exclusive alternatives at
a sequence of times. Consider a model closed system ini-
tially in a pure state that can be described as an observer
and two slit experiment, with appropriate apparatus for
producing the electrons, detecting which slit they passed
through, and measuring their position of arrival on the
screen (Figure 2). Some alternatives for the whole system
are:
1. Whether or not the observer decided to measure
which slit the electron went through.
2. Whether the electron went through the upper or
lower slit.
5
As has been noted by many authors, e.g. [19] and [20] among the
earliest.
FIG. 2: A model closed quantum system containing an ob-
server together with the necessary apparatus for carrying out
a two-slit experiment. Alternatives for the system include
whether the observer measured which slit the electron passed
through or did not, whether the electron passed through the
upper or lower slit, the alternative positions of arrival of the
electron at the screen, the alternative arrival positions regis-
tered by the apparatus, the registration of these in the brain
of the observer, etc., etc., etc. Each exhaustive set of exclusive
alternatives is represented by an exhaustive set of orthogonal
projection operators on the Hilbert space of the closed sys-
tem. Time sequences of such sets of alternatives describe sets
of alternative coarse-grained histories of the closed system.
Quantum theory assigns probabilities to the individual alter-
native histories in such a set when there is negligible quantum
mechanical interference between them, that is, when the set
of histories decoheres. A more rened model might consider
a quantity of matter in a closed box. One could then consider
alternatives such as whether the box contains a two-slit ex-
periment or does not as well as alternative congureatins of
the atoms in the box.
3. The alternative positions, y
1
, , y
N
, that the elec-
tron could have arrived at the screen.
This set of alternatives at a sequence of times denes a
set of histories whose characteristic branching structure
is shown in Figure 3. An individual history in the set
is specied by some particular sequence of alternatives,
e.g. (measured, upper, y
9
).
Many other sets of alternative histories are possible
for the closed system. For example, we could have in-
cluded alternatives describing the readouts of the appa-
ratus that detects the position that the electron arrived
on the screen. If the initial condition corresponded to a
good experiment there should be a high correlation be-
tween these alternatives and the position that the elec-
tron arrives at the screen. In a more rened model we
could discuss alternatives corresponding to thoughts in
the observers brain, or to the individual positions of the
atoms in the apparatus, or to the possibilities that these
atoms reassemble in some completely dierent congura-
tion. There are a vast number of possibilities.
Characteristically the alternatives that are of use to
us as observers are very coarse grained, distinguishing
6
t
t
t
y
y y y
y
LOWER UPPER LOWER UPPER
MEASURED NOT
MEASURED
INITIAL CONDITION
2
3
1
FIG. 3: Branching structure of a set of alternative histories.
This gure illustrates the set of alternative histories of the
closed system illustrated in Figure 2 dened by the alterna-
tives of whether the observer decided to measure or did not
decide to measure which slit the electron went through at
time t1, whether the electron went through the upper slit or
through the lower slit at time t2, and the alternative posi-
tions of arrival y at the screen at time t3. A single branch
corresponding to the alternatives that the measurement was
carried out, the electron went through the upper slit, and ar-
rived at point y9 on the screen is illustrated by the heavy line.
The illustrated set of histories does not decohere because there
is signicant quantum mechanical interference between the
branch where no measurement was carried out and the elec-
tron went through the upper slit and the similar branch where
it went through the lower slit. A related set of histories that
does decohere can be obtained by replacing the alternatives
at time t2 by the following set of three alternatives: (a record
of the decision shows a measurement was initiated and the
electron went through the upper slit); (a record of the de-
cision shows a measurement was initiated and the electron
went through the lower slit); (a record of the decision shows
that the measurement was not initiated). The vanishing of
the interference between the alternative values of the record
and the alternative congurations of apparatus ensures the
decoherence of this set of alternative histories.
only very few of the degrees of freedom of a large closed
system. This is especially true if we recall that our box
with observer and two-slit experiment is only an idealized
model. The most general closed system is the universe
itself, and, as I hope to show, the only realistic closed
systems are of cosmological dimensions. Certainly, as
observers of the universe we utilize only very, very coarse-
grained descriptions of the universe as a whole.
I would now like to state the rules that determine which
coarse-grained sets of histories may be assigned probabil-
ities and what those probabilities are. The essence of the
rules I shall describe can be found in the work of Bob
Griths [17]. The general framework was extended by
Roland Omn`es [18] and was independently, but later, ar-
rived at by Murray Gell-Mann and myself [2]. The idea
is simple: The failure of probability sum rules due to
quantum interference is the obstacle to assigning prob-
abilities. Probabilities can be assigned to just those sets
of alternative histories of a closed system for which there
is negligible interference between the individual histories
in the set as a consequence of the Hamiltonian and par-
ticular initial state the closed system has. All probability
sum rules are satised as a result of the absence of inter-
ference. Let us now give this idea a precise expression.
Sets of alternatives at one moment of time are repre-
sented by sets of orthogonal projection operators. Em-
ploying the Heisenberg picture these can be denoted
{P
k

k
(t
k
)}. The superscript k denotes the set of alter-
natives being considered at time t
k
(for example, the set
of alternative position intervals {y
1
, , y
N
} at which
the electron might arrive at the screen at time t
3
),
k
denotes the particular alternative in the set (for example
y
9
) and t
k
is the time. The set of Ps satisfy

k
P
k

k
(t
k
) = 1 , P
k

k
(t
k
)P
k

k
(t
k
) =

k
P
k

k
(t
k
)
(4.1)
showing that they represent an exhaustive set of exclusive
alternatives.
Sets of alternative histories are dened by giving se-
quences of sets of alternatives at denite moments of
time, e.g. {P
1
1
(t
1
)} , {P
2
2
(t
2
)}, , {P
n
n
(t
n
)}. Dier-
ent choices for {P
1
1
(t
1
)}, {P
2
2
(t
2
)}, etc. describe dier-
ent sets of alternative histories of the closed system. An
individual history in a given set corresponds to a par-
ticular sequence (
1
, ,
n
) and, for each history,
there is a corresponding chain of projection operators
C

P
n
n
(t
n
) P
1
1
(t
1
) . (4.2)
For example, in the two slit experiment in a box illus-
trated in Figure 2, the history in which the observer de-
cided at time t
1
to measure which slit the electron goes
through, in which the electron goes through the upper
slit at time t
2
, and arrives at the screen in position in-
terval y
9
at time t
3
, would be represented by the chain
P
3
y9
(t
3
)P
2
U
(t
2
)P
1
meas
(t
1
) (4.3)
in an obvious notation. The only dierence between this
situation and that of the Copenhagen quantum me-
chanics of measured subsystems is the following: The
sets of operators {P
k

k
(t
k
)} dening alternatives for the
closed system act on the Hilbert space of the closed sys-
tem that includes the variables describing any apparatus,
observers, and anything else. The operators dening al-
ternatives in Copenhagen quantum mechanics act only
on the Hilbert space of the measured subsystem.
When the initial state is pure, it can be resolved into
branches corresponding to the individual members of any
set of alternative histories. The generalization to an im-
pure initial density matrix is not dicult [2], but for
simplicity we shall assume a pure initial state through-
out this article. Denote the initial state by | in the
7
Heisenberg picture. Then
| =

| =

1, ,n
P
n
n
(t
n
) P
1
1
(t
1
)| .
(4.4)
This identity follows by applying the rst of (4.1) to all
the sums over
k
in turn. The vector
C

| (4.5)
is the branch
6
corresponding to the individual history
and (4.4) is the resolution of the initial state into
branches.
When the branches corresponding to a set of alterna-
tive histories are suciently orthogonal the set of histo-
ries is said to decohere. More precisely a set of histories
decoheres when
|C

| 0 , for any

k
=
k
. (4.6)
We shall return to the standard with which decoherence
should be enforced, but rst let us examine its meaning
and consequences.
Decoherence means the absence of quantum mechan-
ical interference between the individual histories of a
coarse-grained set.
7
Probabilities can be assigned to the
individual histories in a decoherent set of alternative his-
tories because decoherence implies the probability sum
rules necessary for a consistent assignment. The proba-
bility of an individual history is
p() = C

|
2
. (4.7)
6
More specically a branch state vector.
7
The term decoherence is used in several dierent ways in the
literature. Therefore, for those familiar with other work, a com-
ment is in order to specify how we are employing the term in this
simplied presentation. We have followed our previous work [2],
[3] in using the term decoherence to refer to a property of a
set of alternative time histories of a closed system. A decoherent
set of histories is one for which the quantum mechanical interfer-
ence between individual histories is small enough to guarantee
an appropriate set of probability sum rules. Dierent notions
of decoherence can be dened by utilizing dierent measures of
interference. The weakest notion is just the consistency of the
probability sum rules that was called consistency by Griths
[17] and Omn`es [18] and that term is used by some to refer to all
measures of interference. Vanishing of the real part of (4.6) is
a sucient condition for the consistency of the probability sum
rules called the weak decoherence condition. We are using the
stronger condition (4.6) because it characterizes widespread and
typical mechanisms of decoherence. Eq (4.6) has been called the
medium decoherence condition. Decoherence in the context
of this paper, thus, means the medium decoherence of sets of
histories. In the literature the term decoherence has also been
used to refer to the decay in time of the o-diagonal elements
of a reduced density matrix dened by tracing the full density
matrix over a given set of variables [21]. The two notions of
decoherence of reduced density matrices and decoherence of
histories are not generally equivalent but also not unconnected
in the sense that in particular models certain physical processes
can ensure both. (See, e.g. the remarks in Section II.6.4 of [22]).
To see how decoherence implies the probability sum
rules, let us consider an example in which there are just
three sets of alternatives at times t
1
, t
2
, and t
3
. A typical
sum rule might be

2
p (
3
,
2
,
1
) = p (
3
,
1
) . (4.8)
We show (4.6) and (4.7) imply (4.8). To do that write
out the left hand side of (4.8) using (4.7) and suppress
the time labels for compactness.

2
p (
3
,
2
,
1
) =

|P
1
1
P
2
2
P
3
3
P
3
3
P
2
2
P
1
1
|
_
.
(4.9)
Decoherence means that the sum on the right hand side
of (4.9) can be written with negligible error as

2
p (
3
,
2
,
1
)

2
2
_
|P
1
1
P
2

2
P
3
3
P
3
3
P
2
2
P
1
1
|
_
. (4.10)
the extra terms in the sum being vanishingly small. But
now, applying the rst of (4.1) we see

2
p (
3
,
2
,
1
)

|P
1
1
P
3
3
P
3
3
P
1
1
|
_
= p (
3
,
1
)
(4.11)
so that the sum rule (4.8) is satised.
Given an initial state | and a Hamiltonian H, one
could, in principle, identify all possible sets of decoher-
ing histories. Among these will be the exactly decohering
sets where the orthogonality of the branches is exact. In-
deed, trivial examples can be supplied by resolving |
into a sum of orthogonal vectors at time t
1
, resolving
those vectors into sums of further vectors such that the
whole set is orthogonal at time t
2
, and so on. However,
such sets of exactly decohering histories will not, in gen-
eral, have a simple description in terms of fundamental
elds nor any connection, for example, with the quasi-
classical realm of familiar experience. For this reason
sets of histories that approximately decohere are of in-
terest. As we will argue in the next two Sections, re-
alistic mechanisms lead to the decoherence of histories
constituting a quasiclassical realm to an excellent ap-
proximation. When the decoherence condition (4.6) is
approximately enforced, the probability sum rules such
as (4.8) will only be approximately obeyed. However,
as discussed earlier, these probabilities for single systems
are meaningful up to the standard they are used. Ap-
proximate probabilities for which the sum rules are sat-
ised to a comparable standard may therefore also be
employed in the process of prediction. When we speak
of approximate decoherence and approximate probabili-
ties we mean decoherence achieved and probability sum
rules satised beyond any standard that might be con-
ceivably contemplated for the accuracy of prediction and
the comparison of theory with experiment.
Decoherent sets of histories of the universe are what
we may utilize in the process of prediction in quantum
8
mechanics, for they may be assigned probabilities. Deco-
herence thus generalizes and replaces the notion of mea-
surement, which served this role in the Copenhagen in-
terpretations. Decoherence is a more precise, more objec-
tive, more observer-independent idea and gives a denite
meaning to Everetts branches. For example, if their as-
sociated histories decohere, we may assign probabilities
to various values of reasonable scale density uctuations
in the early universe whether or not anything like a mea-
surement was carried out on them and certainly whether
or not there was an observer to do it.
V. THE ORIGINS OF DECOHERENCE IN OUR
UNIVERSE
What are the features of coarse-grained sets of histo-
ries that decohere in our universe? In seeking to answer
this question it is important to keep in mind the basic as-
pects of the theoretical framework on which decoherence
depends. Decoherence of a set of alternative histories is
not a property of their operators alone. It depends on the
relations of those operators to the initial state |, the
Hamiltonian H, and the fundamental elds. Given these,
we could, in principle, compute which sets of alternative
histories decohere.
We are not likely to carry out a computation of all
decohering sets of alternative histories for the universe,
described in terms of the fundamental elds, anytime in
the near future, if ever. It is therefore important to inves-
tigate specic mechanisms by which decoherence occurs.
Let us begin with a very simple model due, in its es-
sential features to Joos and Zeh [16]. We consider the
two-slit example again, but this time suppose that in the
neighborhood of the slits there is a gas of photons or
other light particles colliding with the electrons (Figure
4). Physically it is easy to see what happens, the random
uncorrelated collisions can carry away delicate phase cor-
relations between the beams even if the trajectories of the
electrons are not aected much. The interference pattern
will then be destroyed and it will be possible to assign
probabilities to whether the electron went through the
upper slit or the lower slit.
Let us see how this picture in words is given precise
meaning in mathematics. Initially, suppose the state of
the entire system is a state of the electron | > and N
distinguishable photons in states |
1
, |
2
, etc., viz.
| = ||
1
|
2
> |
N
. (5.1)
Suppose further that | is a coherent superposition of a
state in which the electron passes through the upper slit
|U and the lower slit |L. Explicitly:
| = |U +|L . (5.2)
Both states are wave packets in x, so that position in
x recapitulates history in time. We now ask whether
the history where the electron passes through the upper

2
y
FIG. 4: The two slit experiment with an interacting gas. Near
the slits light particles of a gas collide with the electrons. Even
if the collisions do not aect the trajectories of the electrons
very much they can still carry away the phase correlations
between the histories in which the electron arrived at point
y on the screen by passing through the upper slit and that
in which it arrived at the same point by passing through the
lower slit. A coarse graining that described only of these two
alternative histories of the electron would approximately de-
cohere as a consequence of the interactions with the gas given
adequate density, cross-section, etc. Interference is destroyed
and probabilities can be assigned to these alternative histo-
ries of the electron in a way that they could not be if the gas
were not present (cf. Fig. 1). The lost phase information
is still available in correlations between states of the gas and
states of the electron. The alternative histories of the electron
would not decohere in a coarse graining that included both
the histories of the electron and operators that were sensitive
to the correlations between the electrons and the gas.
This model illustrates a widely occuring mechanism by which
certain types of coarse-grained sets of alternative histories de-
cohere in our universe.
slit and arrives at a detector at point y on the screen,
decoheres from that in which it passes through the lower
slit and arrives at point y as a consequence of the initial
condition of this universe. That is, as in Section 4, we
ask whether the two branches
P
y
(t
2
)P
U
(t
1
)| , P
y
(t
2
)P
L
(t
1
)| (5.3)
are nearly orthogonal, the times of the projections be-
ing those for the nearly classical motion in x. We work
this out in the Schrodinger picture where the initial state
evolves, and the projections on the electrons position are
applied to it at the appropriate times.
Collisions occur, but the states |U and |L are left
more or less undisturbed. The states of the photons,
of course, are signicantly aected. If the photons are
dilute enough to be scattered once by the electron in its
time to traverse the gas the two branches (5.3) will be
approximately
P
y
|US
U
|
1
S
U
|
2
S
U
|
N
, (5.4a)
9
and
P
y
|LS
L
|
1
S
L
|
2
S
L
|
N
. (5.4b)
Here, S
U
and S
L
are the scattering matrices from an
electron in the vicinity of the upper slit and the lower slit
respectively. The two branches in (5.4) decohere because
the states of the photons are nearly orthogonal. The
overlap of the branches is proportional to

1
|S

U
S
L
|
1

2
|S

U
S
L
|
2

N
|S

U
S
L
|
N
. (5.5)
Now, the S-matrices for scattering o the upper posi-
tion or the lower position can be connected to that of an
electron at the orgin by a translation
S
U
= exp(ik x
U
)S exp(+ik x
U
) , (5.6a)
S
L
= exp(ik x
L
)S exp(+ik x
L
) . (5.6b)
Here, k is the momentum of a photon, x
U
and x
L
are
the positions of the slits and S is the scattering matrix
from an electron at the origin.
k

|S|k =
(3)
_
kk

_
+
i
2
k
f
_
k, k

k
_
, (5.7)
where f is the scattering amplitude and
k
= |

k|.
Consider the case where initially all the photons are in
plane wave states in an interaction volume V , all having
the same energy , but with random orientations for
their momenta. Suppose further that the energy is low
so that the electron is not much disturbed by a scattering
and low enough so the wavelength is much longer than
the separation between the slits, k|x
U
x
L
| 1. It
is then possible to work out the overlap. The answer
according to Joos and Zeh [16] is
_
1
(k|x
U
x
L
|)
2
8
2
V
2/3

_
N
(5.8)
where is the eective scattering cross section and the
individual terms have been averaged over incoming direc-
tions. Even if is small, as N becomes large this tends
to zero. In this way decoherence becomes a quantitative
phenomenon.
What such models convincingly show is that decoher-
ence is frequent and widespread in the universe for his-
tories of certain kinds of variables. Joos and Zeh calcu-
late that a superposition of two positions of a grain of
dust, 1mm apart, is decohered simply by the scattering
of the cosmic background radiation on the timescale of
a nanosecond. The existence of such mechanisms means
that the only realistic isolated systems are of cosmologi-
cal dimensions. So widespread is this kind of phenomena
with the initial condition and dynamics of our universe,
that we may meaningfully speak of habitually decohering
variables such as the center of mass positions of massive
bodies.
VI. THE COPENHAGEN APPROXIMATION
What is the relation of the familiar Copenhagen quan-
tum mechanics described in Section III to the more gen-
eral post-Everett quantum mechanics of closed systems
described in Sections IV and V? Copenhagen quantum
mechanics predicts the probabilities of the histories of
measured subsystems. Measurement situations may be
described in a closed system that contains both measured
subsystem and measuring apparatus. In a typical mea-
surement situation the values of a variable not normally
decohering become correlated with alternatives of the ap-
paratus that decohere because of its interactions with the
rest of the closed system. The correlation means that the
measured alternatives decohere because the alternatives
of the apparatus with which they are correlated decohere.
The recovery of the Copenhagen rule for when proba-
bilities may be assigned is immediate. Measured quanti-
ties are correlated with decohering histories. Decohering
histories can be assigned probabilities. Thus in the two-
slit experiment (Figure 1), when the electron interacts
with an apparatus that determines which slit it passed
through, it is the decoherence of the alternative congu-
rations of the apparatus that enables probabilities to be
assigned for the electron.
There is nothing incorrect about Copenhagen quan-
tum mechanics. Neither is it, in any sense, opposed to
the post-Everett formulation of the quantum mechanics
of closed systems. It is an approximation to the more gen-
eral framework appropriate in the special cases of mea-
surement situations and when the decoherence of alterna-
tive congurations of the apparatus may be idealized as
exact and instantaneous. However, while measurement
situations imply decoherence, they are only special cases
of decohering histories. Probabilities may be assigned
to alternative positions of the moon and to alternative
values of density uctuations near the big bang in a uni-
verse in which these alternatives decohere, whether or
not they were participants in a measurement situation
and certainly whether or not there was an observer reg-
istering their values.
VII. QUASICLASSICAL REALMS
As observers of the universe, we deal with coarse-
grained histories that reect our own limited sensory per-
ceptions, extended by instruments, communication and
records but in the end characterized by a large amount of
ignorance. Yet, we have the impression that the universe
exhibits a much ner-grained set of histories, indepen-
dent of us, dening an always decohering quasiclassical
realm, to which our senses are adapted, but deal with
only a small part of it. If we are preparing for a jour-
ney into a yet unseen part of the universe, we do not
believe that we need to equip ourselves with spacesuits
having detectors sensitive, say, to coherent superpositions
of position or other unfamiliar quantum variables. We
10
expect that the familiar quasiclassical variables will de-
cohere and be approximately correlated in time by clas-
sical deterministic laws in any new part of the universe
we may visit just as they are here and now.
Since the post-Everett quantum mechanics of closed
systems does not posit a quasiclassical realm, it must
provide an explanation of this manifest fact of everyday
experience. No such explanation can be provided from
the dynamics of quantum theory alone. Rather, like de-
coherence, the existence of a quasiclassical realm in the
universe must be a consequence of both initial condition
of the universe and the Hamiltonain describing evolution.
Roughly speaking, a quasiclassical realm should be a
set of alternative histories that decoheres according to
a realistic principle of decoherence, that is maximally
rened consistent with that notion of decoherence, and
whose individual histories are described largely by alter-
native values of a limited set of quasiclassical variables at
dierent moments of time that exhibit as much as pos-
sible patterns of classical correlation in time. To make
the question of the existence of one or more quasiclassical
realms into a calculable question in quantum cosmology
we need measures of how close a set of histories comes
to constituting a quasiclassical realm. A quasiclassical
realm cannot be a completely ne-grained description for
then it would not decohere. It cannot consist entirely of
a few quasiclassical variables repeated over and over
because sometimes we may measure something highly
quantum mechanical. Quasiclassical variables cannot be
always correlated in time by classical laws because some-
times quantum mechanical phenomena cause deviations
from classical physics. We need measures for maximality
and classicality [2].
It is possible to give crude arguments for the type
of habitually decohering operators we expect to occur
over and over again in a set of histories dening a quasi-
classical realm [2]. Such habitually decohering operators
are called quasiclassical operators. In the earliest in-
stants of the universe the operators dening spacetime
on scales well above the Planck scale emerge from the
quantum fog as quasiclassical. Any theory of the initial
condition that does not imply this is simply inconsis-
tent with observation in a manifest way. A background
spacetime is thus dened and conservation laws arising
from its symmetries have meaning. Then, where there
are suitable conditions of low temperature, density, etc.,
various sorts of hydrodynamic variables may emerge as
quasiclassical operators. These are integrals over suit-
ably small volumes of densities of conserved or nearly
conserved quantities. Examples are densities of energy,
momentum, baryon number, and, in later epochs, nu-
clei, and even chemical species. The sizes of the volumes
are limited above by maximality and are limited below
by classicality because they require sucient inertia
resulting from their approximate conservation to enable
them to resist deviations from predictability caused by
their interactions with one another, by quantum spread-
ing, and by the quantum and statistical uctuations re-
sulting from interactions with the rest of the universe
that accomplish decoherence [23]. Suitable integrals of
densities of approximately conserved quantities are thus
candidates for habitually decohering quasiclassical oper-
ators. These hydrodynamic variables are among the
principal variables of classical physics.
It would be in such ways that the quasiclassical realm
of familiar experience would be an emergent property of
the fundamental description of the universe, not gener-
ally in quantum mechanics, but as a consequence of our
specic initial condition and the Hamiltonian describing
evolution. Whether the universe exhibits a quasiclassi-
cal realm, and, indeed, whether it exhibits more than
one essentially inequivalent realm, thus become calcula-
ble questions in the quantum mechanics of closed sys-
tems.
VIII. CONCLUSION
Quantum mechanics is best and most fundamentally
understood in the context of quantum mechanics of
closed systems, most generally the universe as a whole.
The founders of quantum mechanics were right in point-
ing out that something external to the framework of wave
function and the Schrodinger equation is needed to inter-
pret the theory. But it is not a postulated classical realm
to which quantum mechanics does not apply. Rather it
is the initial condition of the universe that, together with
the action function of the elementary particles and the
throws of the quantum dice since the beginning, is the
likely origin of quasiclassical realm(s) within quantum
theory itself.
Acknowledgments
The formulation of quantum mechanics described in
this paper is a result of joint work with Murray Gell-
Mann. It is a pleasure to thank him for the many con-
versations over the years and for permission to summarize
aspects of our work here. Preparation of the report was
supported in part by the National Science Foundation
under grant PHY90-08502.
[1] In the spirit of providing a simplied introduction to the
quantum mechanics of closed systems rather than a com-
prehensive review, no attempt has been made to provide
anything more than the references that are directly rele-
vant to the points raised in the text as they were in 1992.
Even then these were not always the earliest nor are they
11
the
[2] M. Gell-Mann and J.B. Hartle in Complexity, Entropy,
and the Physics of Information, SFI Studies in the Sci-
ences of Complexity, Vol. VIII, ed. by W. Zurek, Ad-
dison Wesley, Reading; or in Proceedings of the 3
rd
International Symposium on the Foundations of Quan-
tum Mechanics in the Light of New Technology, ed. by
S. Kobayashi, H. Ezawa, Y. Murayama, and S. Nomura,
Physical Society of Japan, Tokyo (1990).
[3] M. Gell-Mann and J.B. Hartle in the Proceedings of the
25th International Conference on High Energy Physics,
Singapore, August, 2-8, 1990, ed. by K.K. Phua and
Y. Yamaguchi (South East Asia Theoretical Physics As-
sociation and Physical Society of Japan) distributed by
World Scientic, Singapore (1990).
[4] S.W. Hawking, Phys. Lett. B 196, 337 (1983).
[5] S. Coleman, Nucl. Phys. B 310, 643 (1988).
[6] S. Giddings and A. Strominger, Nucl. Phys. B 307, 854
(1988).
[7] J. Halliwell, in Quantum Cosmology and Baby Universes:
Proceedings of the 1989 Jerusalem Winter School for
Theoretical Physics, ed. by S. Coleman, J.B. Hartle,
T. Piran, and S. Weinberg, World Scientic, Singapore
(1991) pp. 65-157.
[8] N. Bohr, Atomic Physics and Human Knowledge, Science
Editions, New York (1958).
[9] L. Landau and E. Lifshitz, Quantum Mechanics, Perga-
mon, London (1958).
[10] F. London and E. Bauer, La theorie de lobservation en
mecanique quantique, Hermann, Paris (1939).
[11] H. Everett, Rev. Mod. Phys. 29, 454 (1957).
[12] B. DeWitt and R.N. Graham, eds., The Many Worlds
Interpretation of Quantum Mechanics, Princeton Univer-
sity Press, Princeton (1973).
[13] H.D. Zeh, Found. Phys. 1, 69 (1971).
[14] W. Zurek, Phys. Rev. D 24, 1516 (1981).
[15] W. Zurek, Phys. Rev. D 26, 1862 (1982).
[16] E. Joos and H.D. Zeh, Zeit. Phys. B 59, 223 (1985).
[17] R. Griths, J. Stat. Phys. 36, 219 (1984).
[18] R. Omn`es, Rev. Mod. Phys. 64, 339 (1992).
[19] H.J. Groenewold, Proc. Akad. van Wetenschappen, Am-
sterdam, Ser. B, 55, 219 (1952).
[20] E. Wigner, Am. J. Phys. 31, 6 (1963).
[21] W. Zurek, Physics Today 44, 36 (1991).
[22] J.B. Hartle, The Quantum Mechanics of Cosmology, in
Quantum Cosmology and Baby Universes: Proceedings
of the 1989 Jerusalem Winter School for Theoretical
Physics, ed. by S. Coleman, J.B. Hartle, T. Piran, and
S. Weinberg, World Scientic, Singapore (1991) pp. 65-
157.
[23] M. Gell-Mann and J.B. Hartle, Phys. Rev. D, 47, 3345
(1993).